Boiling cycler nightmare

John Dixon jpcd100 at mole.bio.cam.ac.uk
Wed Aug 16 11:40:49 EST 2000


In article <YE6okCBserm5EAuE at lineone.net>, Dr. Duncan Clark
<Duncan at nospam.demon.co.uk> wrote:

> You will need a very fine thermocouple else the wire actually pulls heat
> from the 50ul of liquid you have in your microtube so it measures wrong.

The thermocouple is in a microtube in the well with the heated lid down
and on 110 deg C. We have left it for several hours on occasion to be
sure the entire set up has equilibrated. 

> You will need a mineral oil overlay as well to stop evaporation. 

50ul of oil's there too - in the thermocouple tube


> It's also not so much (I think) the accuracy of the thermocouple as the
> accuracy of the meter it is attached to. If I remember rightly the
> thermocouple's voltage is constant due to the dissimilar metals and the
> change in voltage per degree and voltage at a set temperature should be
> identical for every manufacturer's thermocouple's. 
> 
> I would test the thermocouple meter calibration in boiling water, ice
> water and how about liquid nitrogen? 
> 

I was trying to think of things that has defined temp between 37 and
100: like something (non-tox and easy to find) that boils at 72 say. 

LN2 is a good one that I hadn't thought of but I'd wouldn't expect it
to be accurate over that large a range of temps and -190 whatever is
not really in the range of temps I am worried about ;-)


> How come the orange dye is getting across the wells? That would suggest
> (apart from possible boiling) that the well seal is leaking? Can you
> eliminate the problem being poor plastics?
> 
 
Only in that the same sealant doesn't allow well jumping in other
cyclers at 99deg C hold. And no noticeable evaporation is seen in other
cyclers, sealed with the same stuff. Plus once we noticed the problem
we pushed on the sealant plastic *really* hard, before running the OG
plate.

On machines we run 9700's and 2400's and touch wood, haven't had
> problems.
> 

Yeah, I think we may be changing manufacturer from now on! No news from
them yet - wonder if they read these groups - hope so :-). 

I have just realised that we are currently complaining to the supplier
not the manufacturer. Might be time to bring them into the loop. They
might be a bit more concerned that their machines are in danger of
getting some fairly severe bad publicity.

> Good luck
> 

Fortunately for me I work furthest from the dodgiest cycler and hardly
ever used it - the guy who works next to it is seriously P***ED off.

> Duncan

-- 
John Dixon                    Lab 44 (1223) 334131
Wellcome/CRC Institute        Fax 44 (1223) 334089
Cambridge University
United Kingdom CB2 1QR       e-m: jpcd100 at mole.bio.cam.ac.uk






More information about the Methods mailing list